Growth Curves and Their Associated Weight and Height Factors in Children
Total Page:16
File Type:pdf, Size:1020Kb
G Model ARCPED-4649; No. of Pages 5 Archives de Pe´diatrie xxx (2018) xxx–xxx Available online at ScienceDirect www.sciencedirect.com Research paper Growth curves and their associated weight and height factors in children from birth to 4 years old in West Azerbaijan Province, northwest Iran a a, b a P. Ghaemmaghami , S.M. Taghi Ayatollahi *, V. Alinejad , Z. Sharafi a Department of Biostatistics, Medical School, Shiraz University of Medical Sciences, Shiraz, Iran b Patient Safety Research Center, Urmia University of Medical Sciences, Urmia, Iran A R T I C L E I N F O A B S T R A C T Article history: Background: Growth trajectory standards are important components that need to be monitored for Received 27 February 2018 suitable child growth. This study examined longitudinal data to identify the factors affecting growth Accepted 30 June 2018 trajectory standards of height and weight for infants. Available online xxx Methods: This study included 566 neonates (286 boys and 280 girls) born in West Azerbaijan Province, northwest Iran, who were followed from birth to 4 years of age. The subjects’ weight and height were Keywords: recorded at birth, 1, 2, 4, 6, 9 months and 1, 1.5, 2, 3, and 4 years of age. In this study, the Lambda-Mu- Growth curve Sigma (LMS) method was used to construct the growth curves for each measure. The linear mixed model Infant was employed to determine the factors affecting the growth trajectory. Weight Height Results: The study demonstrates that the place of birth, duration of breastfeeding, and infants’ gender had LMS method a significant effect on the growth trajectory. Nonetheless, other variables did not have any significant effect on growth. Growth curves for significant factors of weight and height (5th, 50th, 95th percentiles) were obtained. There was a rapid increase in the growth curve from birth to 2 years of age, which then remained relatively constant up to 4 years of age. Discussion: This paper provides the first local growth trajectory standards of height and weight for infants by analyzing longitudinal measurements in West Azerbaijan province. This study determined the factors affecting the growth trend in both indices. It seems that there was a significant difference in the growth trajectories of infants in subgroups of the effective factor. C 2018 Published by Elsevier Masson SAS. 1. Introduction Growth curves, an empirical model to monitor growth patterns over a period of time, are the most widely and acceptable Monitoring growth from birth is a common practice intended to guidelines utilized to evaluate an individual’s health status [6]. This detect growth failure and its association with undesirable factors. curve can be obtained by plotting the increase in size or numbers It is only then that remedial action can be utilized to prevent or against elapsed time [7]. Growth curves are designed based on the reduce nutritional disorders as well as chronic and infectious methods in which the samples are chosen, and also on cross- diseases in childhood, mostly in the first few years of life [1,2]. sectional, longitudinal, and mixed longitudinal design [5,8]. Most Anthropometric measurements as significant indicators for growth studies are based on cross-sectional data to evaluate predicting infants’ health are the most suitable method for growth patterns. Longitudinal studies have also been used to evaluating the nutritional and general health status of a community assess the growth patterns in a few developed countries. These [3,4]. An infant’s weight, height, and head circumference are studies are more valuable when compared to cross-sectional considered to be the most common parameters for investigating an studies, hence, development of the subjects over time can be infant’s physical growth [5]. considered in growth studies [9]. Over the past two decades in Iran, many studies have assessed infant growth utilizing growth curves [5,10–12], but at present no study has assessed the growth standards as a reliable measure for * Corresponding author. a monitoring child growth in West Azerbaijan Province, northwest E-mail address: S.M.Taghi Ayatollahi *[email protected] Iran. (S.M. Taghi Ayatollahi). https://doi.org/10.1016/j.arcped.2018.06.010 C 0929-693X/ 2018 Published by Elsevier Masson SAS. Please cite this article in press as: Ghaemmaghami P, et al. Growth curves and their associated weight and height factors in children from birth to 4 years old in West Azerbaijan Province, northwest Iran. Archives de Pe´diatrie (2018), https://doi.org/10.1016/ j.arcped.2018.06.010 G Model ARCPED-4649; No. of Pages 5 2 P. Ghaemmaghami et al. / Archives de Pe´diatrie xxx (2018) xxx–xxx However, the local standards of one area may differ from those confounder factors, to construct the longitudinal standards for of another. Therefore, our charts can be considered for use as weight and height. The study used the model to verify the standard growth charts in this northwest province because the demographic trend variables over the ordinal variable of time. The samples represent the general features of these areas. However, growth curve model allows the analysis of longitudinal or repeated based on longitudinal data, few studies have attempted to measures data sets where individuals may have missed measu- investigate the influential factors on their growth trajectory rements on one or more occasions. The growth curve model for this [5]. Therefore, this study aimed to present the growth standards study is proposed as: and examine the longitudinal effect of different factors on the growth trajectory measured in boys and girls 0–4 years of age, a birth cohort born in West Azerbaijan province, for weight and y ¼ xb þ zu þ e height, using a linear mixed model. We also compared the growth This is a special case of mixed effects models. In this model, y is charts for weight and height by significant factors. the growth index (weight or height), x is a matrix of fixed effect variables (parents’ age, education, and occupation; child birth order; number of children in the family; duration of breastfeeding; 2. Materials and methods gender; and region (urban or rural)). Z is a matrix of random effects variables (intercept and infants’ age), b and u are parameters of 2.1. Setting and subjects fixed and random effects, respectively. Parameters were estimated utilizing restricted maximum likelihood [14]. A longitudinal study was conducted on the weight and height of The Lambda-Mu-Sigma (LMS) method was used to construct 566 apparently healthy neonates (286 boys and 280 girls), who the growth curves for the measurements of weight and height, were randomly selected by a multistage sampling scheme forming separately for each factor. The LMS is a popular method for a random sample of infants in some health centers from the nine obtaining smooth centile curves [15]. Each centile curve is derived counties (Urmia, Miandoab, Takab, Khoy, Mahabad, Bukan, from three curves representing the median (M), the coefficient of Poldasht, Chaldoran, Salmas) of West Azerbaijan Province in variation (S), and the skewness of distribution (L). Since they 2012. In each county, a random sample of health centers was change with the independent variable (age), the latter is expressed selected and within each health center, using a table of random as a Box-Cox power [16]. The LMS method assumes that skew data numbers, we selected the sample of infants. The subjects were can be normalized by utilizing a power transformation that periodically followed from birth until the age of 4 years at the stretches one tail of the distribution and shrinks the other tail health centers. simultaneously, thus primarily correcting the skewness. Utilizing penalized likelihood, the three curves can be fitted as cubic splines 2.2. Place by nonlinear regression, and the extent of smoothing required can be expressed in terms of smoothing parameters or the equivalent West Azerbaijan Province is one of the 31 provinces of Iran, degree of freedom [3]. located in the northwest of the country, which borders Turkey, and 2 General aspects of the data were analyzed using SPSS version has 27 counties. The province covers an area of 39,487 km , or 2 11.5 (SPSS Inc., Chicago, IL, USA). The growth curve model was 43,660 km including Urmia Lake, with a population of 3 million carried out using SAS 9.2 to determine the significant factors. LMS and an average annual growth rate of 1.4%. Urmia is the capital and chart Maker (professional) and Microsoft Excel software were the largest city of the province. Cold northern winds affect the utilized for the construction of weight and height centiles. The province during winter and cause heavy snow [11]. level of statistical significance was determined as P < 0.05. 2.3. Questionnaire data 3. Results A questionnaire on demographic information and the neonates’ Among the 566 neonates who were followed from birth to 4 years health status of the neonates and their parents including gender; of age within this study, 317 (56%) were urban dwellers. The gender number of children in the family; birth order; parents’ age, education distribution was almost equal, as illustrated in Table 1. The and occupation; age when beginning supplementary food; duration mean Æ SD of the mothers’ and fathers’ age were 25.9 Æ 5.8 and of breastfeeding; region (urban or rural); and anthropometric 30.7 Æ 6.7 years (P < 0.0001), respectively. The mean duration and SD measurements was filled out and analyzed in this study. of breastfeeding was 21.2 Æ 5.85 months. Almost three-quarters of mothers had a normal delivery (420 cases). One hundred fifty-three 2.4. Anthropometric measurements (27.1%) of the families had one child, while 213 (37.7%) had two children.